Draft AI Solution PMI-CPMAI Practice Questions (Page 1)

Questions 1–5

1. 1application · medium

   A manufacturing company wants to implement an AI solution for predictive maintenance on their assembly line equipment. They have sensor data from IoT devices and a legacy CMMS (computerized maintenance management system). The solution must send maintenance alerts to the CMMS. Which integration approach is most appropriate?

   • AUse a message queue like Kafka to publish alerts, and require the CMMS to subscribe to the queue.
   • BUse a direct database connection from the AI system to the CMMS database to insert alerts directly into tables.
   • CUse a REST API to push alerts from the AI system to the CMMS, with JSON payloads containing equipment ID, predicted failure time, and confidence score.
   • DUse a file-based integration where the AI system writes CSV files to a shared network drive that the CMMS polls every hour.
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   Which of the following is a common example of an integration point with a legacy application?

   • AA user interface mockup
   • BA database connection
   • CA model evaluation metric
   • DA training dataset
   Reveal Solution Discussion Select an answer first
4. 3expert · hard

   A logistics company wants to integrate an AI route optimization system with its existing transportation management system (TMS). The TMS is old and only supports SOAP web services. The AI system is new and uses REST APIs. The integration must be reliable and support high message volumes. Which integration approach is most appropriate?

   • AUse an enterprise service bus (ESB) to mediate between the SOAP-based TMS and the REST-based AI system, with message queuing for reliability.
   • BUse a point-to-point integration where the AI system calls the TMS's SOAP service directly, and the TMS calls the AI system's REST API directly.
   • CUse a file-based integration where the AI system exports routes as CSV files and the TMS imports them, with no real-time communication.
   • DModify the TMS to support REST APIs directly, eliminating the need for a middleware.
   Reveal Solution Discussion Select an answer first
5. 4application · medium

   A hospital wants to integrate an AI diagnostic tool with its existing electronic health record (EHR) system. The EHR system is old and only supports HL7 messaging. The AI tool needs to receive patient data and return diagnostic suggestions. What integration approach should be documented?

   • AHave the AI tool poll the EHR system's HL7 feed and parse the messages directly.
   • BDirectly connect the AI tool to the EHR database to read and write patient records.
   • CUse a file-based exchange where the EHR exports patient data as CSV files and the AI tool imports them.
   • DUse an HL7 interface engine to translate between the EHR's HL7 messages and the AI tool's REST API, with JSON payloads.
   Reveal Solution Discussion Select an answer first
6. 5foundation · easy

   Which operational consideration is essential for detecting model performance degradation over time after deployment?

   • AUser interface design
   • BModel monitoring
   • CModel training
   • DData ingestion
   Reveal Solution Discussion Select an answer first